On Fri, 2011-02-18 at 19:03 +0100, Xavier Antoviaque wrote: > Forwarding the answer from Cedric below. I've just replaced the link > to the Twitter Earth demo by an attached screenshot, as it is not > public yet. > > Cedric, what would be the next step for you? Is there something > missing to begin trying something out of this? We could start now, but I prefer to check your models before to be sure there is no something blocking.
Cedric > > Xavier. > > -------- Forwarded Message -------- > From: Cedric Pinson <[email protected]> > Reply-to: [email protected] > To: Xavier Antoviaque <[email protected]> > Subject: Re: [Idea] Less mandatory reading during missions > Date: Fri, 11 Feb 2011 16:34:28 +0100 > > On Thu, 2011-02-10 at 20:12 +0100, Xavier Antoviaque wrote: > > Hi, > > > > We've discussed this a bit yesterday with Cedric (CC'ed), and having a > > first version of the 3D model in WebGL wouldn't be very complicated - > > at least if we start with the following low requirements: > > * No synchronization between the sound played (voice synthesis) > > and the face - the face would be unaltered. Since it is an AI, > > it may not be too disturbing - and it removes lots of > > technical constraints. > > * We would use an existing 3D model, and just contract an artist > > to give the "ethereal" feeling - Cedric knows one which has > > done this type of work > > > > I've done some digging to try to find base models that could fit, and > > I've found two potential candidates: > > * http://www.blendswap.com/3D-models/characters/women-c/ > > * http://www.blendswap.com/3D-models/characters/sintel-v2-5-6/ > > (the main character of the Sintel Blender movie) > > On the captures, the character from Sintel looks very childish, but > > when you load the file the impression disappears (cf attached > > screenshot). It is also more refined. But what do you think? > > > > Also, a quick reminder for Cedric to send the URLs you talked > > about : ) > > * The 3D "twitter" Earth > http://XXXX do not put it in public you will > need chrome 9 or last beta version of firefox4 > > > * The equalizer tests > http://videos.mozilla.org/serv/blizzard/audio-slideshow/ need beta > version of firefox 4 > > > * The voice synthesis API > http://www.oddcast.com/support/NeoSpeech_SSML.html > > Cedric > > > Thanks Cedric! > > > > Xavier. > > > > On Sat, 2011-01-29 at 19:56 +0100, Xavier Antoviaque wrote: > > > Hi, > > > > > > I've been thinking about ways to move some of the content that is > > > currently displayed through the chat window to some other places, to > > > ask less reading from the player. > > > > > > a) There was the option suggested by Nico: filming someone who > > > actually talk to the player. But this option has the huge > > > disadvantage of requiring a lot of work for each mission (actoring, > > > video editing, code manipulations...), and makes it much harder to > > > modify a mission once it has been recorded. > > > > > > b) Deborah recently shown me > > > http://rejecter.blogspot.com/2010/11/so-you-want-to-write-novel_28.html . > > > This wasn't related to the project, but I think it could be a very > > > interesting solution to our problem. We could use text2speech to generate > > > the audio, and put it together with an animation (maybe something like > > > http://code.google.com/creative/radiohead/ ?) > > > > > > I've looked at FS solutions for the text2speech, but in what I've > > > seen the quality isn't good enough yet. Alternatively, we could use > > > services like http://www.ispeech.org/ (Try to type "Good. Here is > > > what I know. Last Monday, the day my agent went off the radar, we > > > received the following message from him:" or "'I am in front of this > > > painting, the only one that was ever sold by a man who said once : > > > "). We could generate the MP3 there and distribute them, if their > > > license allows it (to check). It would be as proprietary as a real > > > voice anyway... > > > > > > c) Another lead, which is complementary, is the idea David had to > > > have some kind of "Mission file" - where the mission goals and steps > > > would be explained, and where you could find some content (images, > > > video...) related to the mission. This could help to shorten the > > > text, which would likely still be an issue even with a voice or a > > > video. > > > > > > At the end of the day, in any case this is a significant work - but > > > the result could be worth it. What do you think? If we try some of > > > those options, who would like to work on this? > > > > > > Idea on the feedback tool: > > > http://feedback.farsides.com/forums/90557-humanity-project-accessibility/suggestions/1427705-replace-the-chat-by-text2speech-ai-vizualisation?ref=title > > > > > > > > > Xavier. > > > _______________________________________________ > > > Farsides mailing list - [email protected] > > > > > > Wiki: http://farsides.com/ > > > List: http://farsides.com/ml/ > > > Forum: http://farsides.com/forum/ > > > Ideas: http://farsides.com/ideas/ > > > Chat: http://farsides.com/chat/ > > > > -- > Provide OpenGL, WebGL and OpenSceneGraph services > +33 659 598 614 Cedric Pinson mailto:[email protected] > http://www.plopbyte.net > -- Provide OpenGL, WebGL and OpenSceneGraph services +33 659 598 614 Cedric Pinson mailto:[email protected] http://www.plopbyte.net
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Farsides mailing list - [email protected] Wiki: http://farsides.com/ List: http://farsides.com/ml/ Forum: http://farsides.com/forum/ Ideas: http://farsides.com/ideas/ Chat: http://farsides.com/chat/
